Articles by Hassam Nasir

An update by Intel to the perfmon platform has added support for upcoming Panther Lake CPUs, listing their core architecture codenames and CPUID, via InstLatX64. This commit unofficially confirms that Panther Lake will employ Cougar Cove Performance (P) cores, while Darkmont will serve to power its Efficiency (E), and likely Low Power Efficiency (LPE) cores as well. Panther Lake is expected to launch later this year, succeeding current-generation Arrow Lake U/H offerings.With Intel’s flagship 18A in risk production, Panther Lake is scheduled for mass production later this year. Hence, it wouldn’t be surprising if the bulk of Panther Lake arrives in Q1 next year, similar to how Meteor Lake rolled out. Make no mistake: Panther Lake isn’t a successor to Lunar Lake, which was uniquely focused on efficiency as a one-off product, with on-package memory, limited TDP, and a power-optimized design. Current rumors indicate Panther Lake variants will sport up to 18 hybrid cores (6P+8E+4LPE) and 12 Xe cores, based on Intel’s upcoming Celestial (Xe3) graphics architecture.Intel positions Panther Lake as combining Arrow Lake’s power and Lunar Lake’s efficiency, but that’s still a somewhat general claim. According to leaks, most Panther Lake systems will include traditional SODIMM/soldered memory, while some laptop designs might even support next-gen LPCAMM, combining fast and upgradeable RAM. Based on their TDP (rumored: up to 64W), Panther Lake chips are expected to power a wide range of devices, including entry-level laptops, handhelds, and gaming laptops. The company is even eying bringing this architecture to automobiles.

You may like

An Intel engineer has pushed an update to the lookup table for perfmon, adding Panther Lake as a supported architecture. Panther Lake has been marked with the “GenuineIntel-6-CC” identifier, assigning it to CPU Family 6, Model 204 (0xCC). In addition, the patch reports the Cougar Cove and Darkmont architectures for Panther Lake’s Performance (P) and Efficiency (E) cores, respectively.#Intel perfmon: #PTL=CougarCove+DarkmontCPUID.1Ah.EAX =20000000 #Tremont20000001 #Gracemont20000002 #Crestmont20000003 #Skymont20000004 #Darkmont40000000 #SunnyCove40000001 #GoldenCove40000002 #RedwoodCove40000003 #LionCove40000004 #CougarCovehttps://t.co/adntAfZRm5 https://t.co/1in8Rvleue pic.twitter.com/bWZyYtZcViApril 27, 2025Intel’s speculated P-core evolution goes from Lion Cove (ARL/LNL) to Cougar Cove (PTL/WCL), and then Coyote Cove (NVL). Likewise, the rumored E-core path is Skymont (ARL/LNL) to Darkmont (PTL/WCL) and Arctic Wolf (NVL). Once again, these are based on preliminary assumptions and a compilation of several leaks, so take them with a heavy dose of skepticism.Essentially, we’re seeing a full architectural jump and shift to a leading-edge node, which should yield substantial performance gains. I’m curious to see how much Panther Lake laptops will cost compared to Lunar Lake, which was entirely manufactured by TSMC. With Computex just weeks away, we expect Intel to share finer architectural details at the show.Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox. […]

Like Bethesda’s Oblivion Remastered, it looks like we’re getting another shadow drop this month, this time from AMD. Pre-orders for the Radeon RX 9070 GRE are now live at major e-commerce retailers in China, with deliveries expected to begin on May 8. Adding the ‘officially confirmed’ stamp to this news, the product page for the RX 9070 GRE has gone live on AMD’s website, detailing all the specifications.The RX 9070 GRE has been generating buzz in the news since early April. Consistent with AMD’s past GRE releases, it was almost confirmed that the RX 9070 GRE would be region locked to China, at least initially. While the RX 7900 GRE set a precedent for later global launches, that was largely to rival Nvidia’s RTX 4070 Super. The RX 9070 GRE sits between the RX 9070 and the yet-to-launch RX 9060 XT. That’s what we would naturally assume; however, a price almost identical to the faster RX 9070 hints AMD might be replacing the latter with the GRE, at least in China.In terms of specifications, the RX 9070 GRE wields 48 Compute Units, for a total of 3,072 Stream Processors, 25% fewer than the RX 9070 XT. It likely sticks to the same Navi 48 core as seen in the RX 9070 series, along with a 192-bit interface for 12GB of 18 Gbps GDDR6 VRAM. The boost clocks are up there at 2.79 GHz, right between the RX 9070 and RX 9070 XT. The GPU has a 220W TGP, matching RX 9070, and AMD recommends a 650W PSU for it. Most models should be powered by 2x 8-pin connectors, and some AIBs might even carry over existing RX 9070 board designs due to the similar power specifications.

You may like

Swipe to scroll horizontallySpecificationRX 9070 GRERX 9070RX 9070 XTRX 7900 GREArchitectureRDNA 4RDNA 4RDNA 4RDNA 3Compute Units (CUs)48566480Stream Processors3,0723,5844,0965,120Game Clock 2.22 GHz2.07 GHz2.40 GHz1.29 GHzBoost Clock 2.79 GHz2.52 GHz2.97 GHz2.25 GHzMemory12GB GDDR616GB GDDR616GB GDDR616GB GDDR6Memory Bus192-bit256-bit256-bit256-bitMemory Speed18 Gbps20 Gbps20 Gbps18 GbpsMemory Bandwidth432 GB/s640 GB/s640 GB/s576 GB/sTDP (approx.)220W220W304W260WPower Connectors2x 8-pin2x 8-pin2x/3x 8-pin2x 8-pinInterfacePCIe 5.0 x16PCIe 5.0 x16PCIe 5.0 x16PCIe 4.0 x16Launch Price (China)4,199 RMB ($575)4,499 RMB ($600) / $549 Global4,999 RMB ($685) / $599 USD (Global)5,299 RMB ($700) at launch in China / $549 USD (Global)The slower GDDR6 modules and narrower 192-bit interface chop the memory bandwidth by 32.5% compared to the rest of the lineup. Depending on how RDNA 4 scales with bandwidth, this might slightly impact performance. As seen before, there’s a decent chance you can extract some gains by overclocking the memory, but let’s wait for official tests.AMD, through its first-party metrics, is claiming 6% better performance than the RX 7900 GRE on average across a set of 30 games at 1440p Ultra. That’s a lot better than what we initially predicted, and underlines RDNA 4’s per Compute Unit improvements in rasterization. But we’ll need to see third-party/independent reviews before jumping to conclusions.AIBs are commanding anywhere between 4,199 RMB ($575) to 4,499 RMB ($620), so we’ll assume the former as AMD’s recommended MSRP. Without China’s general 13% VAT, the price comes down to roughly $500. While this might seem high at first glance, the RX 7900 GRE also launched at 5,299 RMB ($700) in China, followed later by a $549 global release.It’s possible that either the GRE is replacing the standard RX 9070 in China, or local prices for the GRE will drop gradually. Only time will tell, but we’re currently awaiting reviews from Chinese enthusiasts and hoping a handful of these cards slip onto U.S. shores.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox.Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button. […]

Earlier this month, reports suggested that Nvidia has added SK hynix as a GDDR7 supplier for its RTX 50 series GPUs, after previously only sourcing from Samsung. On Chiphell, a user recently got their hands on a Gigabyte Gaming OC RTX 5070 Ti with SK hynix memory and found its overclocking potential similar to what we’ve seen from Samsung, via UNIKO’s Hardware. While thermal and efficiency data are missing, this is nonetheless a positive indication.Each generation, Nvidia casts a wide net to broaden its memory supplier base, ensuring multiple options. This is a common approach with many products, especially SSDs, many of which often ship with revised controllers or NAND flash modules later in their lifecycle. This shouldn’t impact the average consumer as Nvidia likely validates all memory chips to run at a minimum of 28 Gbps (or 30 Gbps for the RTX 5080), at defined power, voltage, and temperature settings.A user at Chiphell secured an RTX 5070 Ti from Gigabyte (SK hynix memory) earlier this week. Several days later, they attempted to increase the power limits by flashing a BIOS that was designated for the Aorus Master RTX 5070 Ti (Samsung memory), resultantly bricking the card. Fortunately, the damage wasn’t permanent as their GPU was equipped with a dual BIOS, allowing them to switch to the secondary firmware.

You may like

Either way, to test the limits of their GPU’s memory, the user achieved a memory clock of 2,125 MHz (34 Gbps) with overclocking, aligning with the majority of Samsung overclocks we’ve witnessed. This tells us that GDDR7 modules from SK hynix are just as capable as ones from Samsung when it comes to overclocking – with this sample.hynix g7 oc tested by user michelelee.hynix g7 on 5070ti runs at 28gbps out of the box and it can be overclocked to 34gbps without problems.although bios for Samsung g7 can be flashed on cards that use hynix g7, the card will be bricked.user michelelee on chh shares his… https://t.co/T6Yn7LjL3J pic.twitter.com/77ZLbYJRfnApril 26, 2025A common theory being put forward is that RTX 50 BIOS versions are memory vendor-dependent, so firmware for Samsung will not work on SK hynix-based cards, and conversely. However, it’s difficult to draw a reasonable conclusion as our test data is limited to just one example.Even though memory offerings from SK hynix, Micron, and Samsung are mostly comparable and rigorously validated by Nvidia, there have been slight inconsistencies in the past. Likewise, the overclocking headroom, voltage requirements for specific speeds, and thermals can differ between memory technologies, making it a bit of a lottery for enthusiasts and overclockers.Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox. […]

For over a year, hardware enthusiast Daniel Simionescu (sdz) has been hard at work porting 3dfx’s (now-defunct) Voodoo graphics to the mobile form factor. Now, the fruits of their efforts have materialized in the form of several designs, with one including a Voodoo4 GPU integrated into a Dell Precision M4800 laptop using a custom-designed MXM (Mobile PCI Express Module) card. YouTuber Omores got their hands on this exact setup and decided to take it for a spin in Windows 98 and XP. They encountered several compatibility issues, but eventually got the GPU up and running, at least partially.

World’s First Laptop with 3dfx Voodoo 4, but Windows 98 Chooses iGPU — Had to Use VBE Drivers – YouTube

Watch On
Voodoo was all the hype in the late 90s, being among the first affordable and widely adopted 3D accelerators for the mainstream. As we mark the 25th anniversary of Nvidia’s takeover of 3dfx, Voodoo accelerators are now mostly remembered for nostalgia. At the Vogons forums, dedicated modder sdz developed a PCB equipped with a Voodoo4 GPU, specifically the VSA-100 core paired with 64MB of SDR memory, mirroring the never-released Voodoo4 4800.Bridging the gap between the VSA-100 and modern standards required the integration of an FPGA, an LVDS/eDP connector, and a RealTek Scaler on the carrier board. YouTuber Omores decided to put this design to the test with the Voodoo4 board connected to a Dell Precision M4800, chosen for its Haswell-based CPU and, crucially, its MXM slot.

You may like

Testing in Windows 98 yielded no significant results, as the BIOS flat-out refuses to use Voodoo4 as the primary GPU during boot-up. Likewise, the Voodoo4’s BIOS isn’t MXM-compliant, and the laptop’s BIOS uses a whitelist, only allowing use with authorized cards. As a workaround, the YouTuber had to switch to generic VBE (VESA BIOS Extensions) drivers, which offer higher display resolutions and a basic display output with little to no hardware acceleration. The video goes over several VBE drivers, including VBESVGA, VB9X, and Scitech Display Doctor, each with its own set of limitations and quirks.That being said, Windows XP cooperated slightly better, and the YouTuber was able to fire up the 3DMark 2001 benchmark and Need for Speed: Porsche Unleashed, without any major hiccups. Running games built using 3dfx’s proprietary Glide API on this system might require using wrappers. Regardless, there isn’t a definitive list of supported games, but we’re sure the community might compile one soon enough. The original developer doesn’t have any intention to commercialize this design, but instead plans to make it open-source.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox. […]

Where most would buy a dedicated water block for their processor, crafty Chinese YouTuber octppus machined a cold plate directly from their CPU’s IHS, via UNIKO’s Hardware at X. Testing this approach to yield tangible data required some rather unwieldy strategies, like drawing water from a bucket to ensure a controlled environment. While the idea of carving a water block from your IHS seems ingenious, it carries the obvious risk of destroying your CPU and presents some quirky thermal characteristics.Why would you even do this in the first place? Your CPU’s die is connected to the IHS (Integrated Heat Spreader) by STIM or thermal grease. To liquid cool your CPU, a metal block (or cold plate) is essential to conduct heat from the IHS. Water blocks contain intricate microchannels or streams, where the heat from the block’s metal is transferred to the coolant via convection. Since each interface or surface increases thermal resistance, serious overclockers typically remove the IHS and mount the water block directly onto the die.What if you took this the other way around, and removed the external waterblock instead? With the help of CNC machines, in collaboration with a viewer, the modder was able to carve a series of interconnected channels or streams in the IHS of their Core i9-14900KS for the coolant. There are also visible areas for the inlet and outlet fittings, and the entire structure is enclosed in a screwed-in gasket to ensure a leak-proof design.

You may like

Image […]

As revealed by Phoronix, Intel’s new ‘200S Boost’ feature for its Arrow Lake chips results in little to no performance improvements on Linux. This parallels our 200S Boost testingin Windows, where memory overclocking from DDR5-6400 to DDR5-8000 accounted for most of the performance increases across gaming and productivity.In the six months following Arrow Lake’s debut, Intel has issued several fixes to extract every last bit of performance from these chips. The first wave was delivered via Windows Updates and updated BIOS versions in December, addressing several major issues. Following that, board partners released the long-awaited 0x114 microcode in January through BIOS updates carrying the ME19.0.0.1854v2.2 firmware.Earlier this month, Intel debuted its IPO program in China, allowing System Integrators to offer enhanced power ratings and clock speeds under warranty. Similar to IPO, Intel’s ‘200S Boost’ profiles are opt-in BIOS presets available globally, enabling higher fabric, die-to-die, and memory-transfer speeds than stock.

You may like

Phoronix’s test suite features the Core Ultra 9 285K, the Asus ROG Maximus Z890 Hero (1801 BIOS), the RX 7900 XTX, and 32GB of DDR5-6400 memory. The tests were conducted in Ubuntu 25.04 with the Linux kernel version 6.14. Gaming remained largely similar to stock, with small improvements in Counter-Strike 2 and Batman: Arkham Knight, while Dirt Rally 2.0 saw a significant drop in FPS (456 FPS vs 406 FPS) when switching to the boost profile.Moving to productivity, compilation tests were slightly faster with no real impact on rendering workloads. That’s generally the sentiment across most real-world or synthetic benchmarks, and there’s a good reason for that. The most significant factor driving the gains observed with Intel’s 200S Boost is the support for faster memory speeds. If you want to get the most out of these profiles, consider faster DDR5-7200 or DDR5-8000 memory kits, because the results from only overclocking the NGU or D2D interconnects are imperceptible. Our testing corroborates that.Phoronix’s test bench stuck to a standard DDR5-6400 kit, and the improvements were so tiny you needed to look for them. We saw a similar effect in Windows when comparing the boost profile vs stock using the same DDR5-6400 kit, for a 0.8% increase in FPS across a geomean of 16 games.While this is a step in the right direction, you’ll need a better memory kit for actual improvements. Likewise, these profiles don’t exactly guarantee stability with faster RAM. They only offer warranty coverage, and I’m guessing most people who buy DDR5-8000 kits are already using them at their rated speeds with XMP.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox.Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button. […]

The Ryzen 5 7533HS is a new Rembrandt-R-based APU that AMD quietly introduced to the market last year and is just now appearing in budget laptops via realVictor_M at X. To clarify, this isn’t exactly a brand-new chip; it is simply a rebadged Ryzen 5 7535HS, with lower boost clocks and presumably a more attractive price tag for partner(s).The Ryzen 5 7533HS breaks away from AMD’s original naming convention for the Ryzen 7000/8000 series, where the last digit was used to indicate versions within an architecture (‘0’ for the lower model and ‘5’ for the upper model), like Zen 3 (7530) versus Zen 3+ (7535). This isn’t entirely surprising as AMD has a history of abandoning naming schemes, evident with their latest shift to “Ryzen AI”. The Ryzen 5 7533HS is tagged under the Rembrandt-R family, with Zen 3+ cores and an RDNA 2-based iGPU.In terms of specifications, the Ryzen 5 7533HS is nothing special. It wields a six-core / twelve-thread layout, along with 3MB of L2 and 16MB of L3 cache. The included Radeon 660M iGPU offers six RDNA2-based Compute Units. There’s a 150 MHz drop in boost clocks compared to the Ryzen 5 7535HS. As the ‘HS’ modifier indicates, the APU has a configurable TDP between 35W and 54W, with support for DDR5-4800 (SODIMM) and LPDDR5-6400 (Soldered RAM) memory types.

You may like

Swipe to scroll horizontallyCPURyzen 7 7735HSRyzen 5 7535HSRyzen 5 7533HSFamilyRembrandt-RRembrandt-RRembrandt-RCores/Threads8/166/126/12iGPU ModelRadeon 680MRadeon 660MRadeon 660MCompute Units1266Base Clocks3.20 GHz3.30 GHz3.30 GHzBoost Clocks4.75 GHz4.55 GHz4.40 GHzL2/L3 Cache4MB/16MB3MB/16MB3MB/16MBTDP35W-54W35W-54W35W-54WWhile the Ryzen 5 7533HS won’t certainly top any performance charts, it still appears to be a solid choice for affordable laptops. On that note, this chip has only appeared in several of Lenovo’s newest laptops, like the ThinkBook 16 Gen 7, IdeaPad Slim 3/5, just to name a few. This strongly hints at Lenovo exclusivity, and that’s not surprising considering that rumors suggest AMD’s latest Ryzen Z2 Go, launched in January, was developed explicitly for Lenovo’s Legion Go S.As it stands, AMD’s mobile portfolio looks something like this (in order of descending price and performance): Strix Halo (Ryzen AI Max), Strix Point (Ryzen AI 360/370), Krackan Point (Ryzen AI 340/350), Hawk Point (Ryzen 200), and Rembrandt-R (Ryzen 7035). There might be some shuffling with a potential Gorgon Point (Strix Point Refresh), but don’t expect any massive changes until Medusa Point (Zen 6 APUs).Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox. […]

Motherboard vendors have started to deploy BIOS updates based on the AGESA 1.2.0.3C firmware. The new BIOS addresses a critical security vulnerability in AMD’s Zen 5 chips found last month. This security flaw impacts Zen-based microprocessors across all product lines. While firmware updates patched Zen 1 to Zen 4, this vulnerability was only recently discovered with Zen 5.According to AMD’s security bulletin, the company relayed the updated firmware to motherboard vendors late last month. Due to the time each partner needs to integrate and validate new firmware for their unique BIOS on each motherboard model, we’re only starting to see adoption now. So far, only MSI has the updated BIOS for some of its 800-series motherboards.The specific vulnerability in question is called EntrySign (ID: AMD-SB-7033), and it allows unsigned and potentially malicious microcode to be executed on the CPU. The flaw stems from AMD’s signature verification process, which used a weak hashing algorithm (AES-CMAC). This allowed researchers at Google to craft forged signatures for arbitrary or even malicious microcode. The catch is that said bad actors must have kernel-level (ring 0) privileges, and at that point, this bug should be the least of your concerns, at least in consumer-grade environments.

You may like

To be clear, hot-loaded microcodes don’t persist across reboots. Every time you power down and reboot your system, the microcode resets to one that was permanently embedded in your CPU from the factory unless changed later in the boot process by the BIOS/OS, which adds another set of guardrails.MSI releases AGESA ComboAM5PI 1.2.0.3C BIOS for its AM5 800-series motherboards, to fix “AMD Microcode Signature Verification Vulnerability” (CVE-2024-36347) reported by researchers from Google. 🧐🧐🧐 pic.twitter.com/T8ky1mBRR5April 25, 2025In addition to desktops, this vulnerability posed a risk to server-grade processors like AMD’s Turin (EPYC 9005) family, potentially compromising their SEV and SEV-SNP protection technologies, which could allow unauthorized access to private data from virtual machines. As it stands, except for Fire Range (Ryzen 9000HX), mitigation is available for all CPUs from the Zen 5 family: Granite Ridge, Turin, Strix Point, Krackan Point, and Strix Halo.This has several implications for the average user. A typical example can be BYOVD (Bring Your Own Vulnerable Driver) attacks, where hackers abuse vulnerabilities in trusted and signed kernel-level drivers to gain access to ring 0. If successful, this could be a stepping-stone to exploiting CPU vulnerabilities like EntrySign, allowing them to execute malicious microcode on your processor.An example of this is when hackers discovered holes in Genshin Impact’s (kernel-level) anti-cheat and distributed ransomware that targeted this flaw and achieved ring 0 access. In short, the safest approach is to keep an eye out for an upcoming BIOS update from your vendor that has been specified to use the AGESA 1.2.0.3C firmware.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox.Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button. […]

Nvidia’s budget-friendly RTX 5060 GPUs are rumored to hit shelves on May 19th, according to VideoCardz’s sources. This release window falls just a day before Computex, and a day after AMD’s RX 9060 XT series, allegedly. Nvidia is reportedly relaying the embargo timelines to AIB partners as we speak, which is likely the origin of this leak. Since embargoes are always subject to change, we should treat this leak with skepticism.Last week, Nvidia unveiled the RTX 5060 family of Blackwell GPUs, including the RTX 5060 Ti 8GB/16GB ($429/$379) and the RTX 5060 ($299), with the latter slated for an unspecified launch date next month. Contrary to a typical one-day gap between the review and sales embargoes, VideoCardz reports that for the RTX 5060, both will be lifted on the same day: May 19th. Instead of relying on Nvidia’s skewed first-party benchmarks, interested customers should wait for independent reviews before jumping the gun.With the sharp spike in shader count and bandwidth, you could be looking at 20-30% higher performance than the RTX 4060, but that’s just a rough estimate. Equipped with the GB206 core, the RTX 5060 features 3,840 CUDA cores, 25% more than the RTX 4060, along with a 128-bit interface for 8GB of memory. Thanks to the adoption of GDDR7, there is a 64% increase in memory bandwidth (from 288 GB/s to 448 GB/s), which will prove useful in memory-intensive tasks.

You may like

The RTX 5060 Ti 16GB offers notable improvements over its predecessor, thanks to increased memory bandwidth and architectural refinements, and is otherwise a solid choice if you can obtain it at MSRP. Despite sharing the same core, its 8GB brethren suffers greatly in VRAM-bound titles, so you’ll have to acknowledge its limits.With a potential RTX 5050 on the horizon, Nvidia’s Blackwell lineup might include three GPUs sharing the same, limited 8GB framebuffer. This is obviously to segment its offerings, forcing users to invest over $400 should they want more than 8GB of VRAM, which is borderline essential for 1440p gaming. Testing shows that modern AAA titles can be so taxing at the highest settings that they can completely saturate an 8GB framebuffer, even when playing at 1080p.The RTX 5060 might shape up to be a great pick for budget-minded gamers aiming for a modern GPU without emptying their wallets. However, you should be aware of its limitations. That being said, all eyes are on availability, as RTX 50-series GPUs are practically a unicorn at MSRP.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox. […]

Without any flash or pomp, Kingston’s PCIe 5.0-ready Renegade G5 SSDs are now available for purchase, starting at $200. ComputerBase first reported on the listings. The new flagship drive boasts read speeds of up to 14.8 GB/s, rivaling Samsung’s fastest 9100 Pro offering, but edges it out in sequential writes. The SSD is officially listed on Kingston’s website, along with a detailed datasheet, and you can snag one for yourself directly through their store or other resellers.This year, the PCIe 5.0 arena has been crowded with new contenders all vying for the speed crown. High-end PCIe 5.0 SSDs were ruled by predominantly ruled by Phison’s E26 controller, until Silicon Motion’s SM2508 arrived last year to challenge its reign.Kingston’s latest Renegade G5 appears to be among the first few implementations of the Silicon Motion controller, and on-paper specifications are certainly promising. It is somewhat peculiar that Kingston opted for a low-key announcement of its flagship SSD, superseding the in-vogue KC3000 and Fury Renegade PCIe 4.0.

You may like

Kingston has equipped the Renegade G5 with TLC NAND flash, which we presume is based on Micron’s 232-layer technology. It comes loaded with DRAM cache, but the exact capacity has not been specified.The G5 can hit sequential read and write speeds of up to 14.8 GB/s and 14 GB/s, surpassing the best from Samsung and Crucial. However, in random write IOPS, the strength of Samsung’s Presto controller is evident with the 9100 Pro achieving 2,600K IOPS, which is higher than the G5’s 2,200K IOPS.Swipe to scroll horizontallyNameKingston Renegade G5Samsung 9100 ProCrucial T705Flash MemoryTLC (Likely 232-Layer NAND flash from Micron)236-Layer Samsung TLC NAND232-Layer Micron TLC NANDForm FactorM.2 2280M.2 2280M.2 2280ControllerSM2508Samsung Proprietary (Presto)Phison E26DRAMYesYesYesTBW (for 4TB variant)4,000TB2,400TB2,400TBSequential Reads14.8 GB/s14.8 GB/s14.5 GB/sSequential Writes14.0 GB/s13.4 GB/s12.7 GB/sRandom Reads2,200K2,200K1,550KRandom Writes2,200K2,600K1,800KWhen it comes to TBW, which indicates the amount of data that’s guaranteed to be written on a drive, the Renegade G5 stands out with a 4,000TB endurance rating (4TB variant), significantly higher than the 2,400TB offered by its competitors.The maximum rated power consumption ranges from 6.6W (1TB) to 9.5W (4TB), slightly better than some implementations of the Phison E26. Either way, we’ll have to double-check these numbers in independent testing.Get Tom’s Hardware’s best news and in-depth reviews, straight to your inbox.The Renegade G5 is available in 1TB, 2TB, and 4TB capacities at $200, $329, and $629, respectively. Sadly, unlike the 9100 Pro, there isn’t a fully decked-out 8TB variant available yet. […]