Chinese Vendor Designs PCIe 4.0 GPU, Targets GTX 1080 Performance


This site may earn affiliate commissions from the links on this page. Terms of use.

The high-performance GPU industry has been a two-horse race for very nearly two decades. After the collapse of 3dfx, no new company emerged to seriously challenge the ATI/Nvidia split. While Intel holds a substantial stake of the total GPU market, its integrated business has only focused on 2D, video, and basic gaming 3D. Intel’s upcoming Xe architecture, expected in 2020, will take a serious shot at breaking into the consumer space. Now, there’s a word of a potential fourth player in the field, albeit it possibly in a more specialized area.

According to THG, Jingjia Micro is a military-civilian integrated company that’s primarily focused on developing GPUsSEEAMAZON_ET_135 See Amazon ET commerce for the military market thus far. The company began by building China’s first homegrown GPU, the JM5400, built on 65nm. The success of the JM5400 allowed the company to expand and move to newer manufacturing nodes. Its next products, the JM7000 and 7200, were built on 28nm. Now, Jingjia Micro wants to expand its reach further and target the performance of the GTX 1050 and 1080 with a pair of new designs — the JM9231 and JM9271.

Jingjia-Micro-Chart

Chart by THG

A post at cnbeta has additional information. Currently, the JM7200 is said to offer performance equivalent to the GeForce GT 640, albeit in a much lower power envelope — 10W, supposedly, compared with the 50W Nvidia specced for that card. We’d like to see that claim independently verified. The OEM variant of the GT 640 was a Fermi-based part built on 40nm, but that chip had a 65W TDP. The 50W variant was a Kepler-derived part built on 28nm — the same process node Jingjia Micro uses. The JM part also supposedly has 4GB of RAM, while the GT 640 50W version had just 1GB of GDDR5.

The JM9231 and JM9271 are supposedly the first fully programmable GPUs that Jingjia Micro has developed; there are references to the previous JM5400 and JM7200 families being based on fixed-function rendering pipelines. These limitations wouldn’t fly under modern APIs for Windows, but the company started life as a military GPU vendor, and such applications obviously have very different requirements for APIs and product certification.

The new JM parts obviously aren’t going to gun for the highest-end cards from Nvidia or AMD, but even approaching high-end performance from 2016 – 2017 would allow them to contend for the midrange and budget markets. Bringing up the software stack and winning developer support would obviously be critical to any market play, and there doesn’t seem to be any information about whether the JM9231 or JM9271 include any performance improvements or ideas that we haven’t seen before from the major vendors. Such events are rare, but not unheard of. PowerVR once attempted to establish itself as a third player in PC graphics with the Kyro and Kyro II, which won some market share for itself as a unique solution with higher memory bandwidth efficiency than either ATI or Nvidia.

The use of HBM memory in a product of this sort is rather interesting, as is the comparatively low memory bandwidth (by HBM standards). Given that both products lack modern API support, it’s possible they’re intended strictly for military use — though in that case, referencing the GTX 1080 would be a bit odd. Either way, China clearly has its eye on competing more aggressively in terms of overall silicon performance. A few more years, and we might see new products from vendors we haven’t seen before challenging ‘homegrown’ alternatives like AMD, Nvidia, and (if its Xe launch goes well), Intel.

Now Read: 




10 minutes mail – Also known by names like : 10minemail, 10minutemail, 10mins email, mail 10 minutes, 10 minute e-mail, 10min mail, 10minute email or 10 minute temporary email. 10 minute email address is a disposable temporary email that self-destructed after a 10 minutes. https://tempemail.co/– is most advanced throwaway email service that helps you avoid spam and stay safe. Try tempemail and you can view content, post comments or download something

Why 110-Degree Temps Are Normal for AMD’s Radeon 5700, 5700 XT


This site may earn affiliate commissions from the links on this page. Terms of use.

AMD has published a blog post discussing how temperatures and thermals are calculated on its Navi GPUs. There has been some concern in the enthusiast community about the temperatures posted by reference cards, given that these GPUs can report thermal junction temps of up to 110 degrees Celsius. This is substantially hotter than the old temperature of 95 C, which used to be treated as a thermal trip point.

Beginning with Radeon VII, AMD made significant changes to how it measures temperature across the GPU die. In the past, AMD writes, “the GPU core temperature was read by a single sensor that was placed in the vicinity of the legacy thermal diode.” That single reading was used to make decisions governing the GPUs voltage and operating frequency. Radeon VII and now Navi do things differently. Instead of deploying a single sensor, they use a network of sensor data gathered from across the GPU. AMD has deployed the same AVFS (Adaptive Voltage and Frequency Scaling) strategy that it uses for Ryzen to maximize performance of its GPUs.

AVFS deploys a network of on-die sensors across the entire chip rather than relying on a single point of measurement. Rather than calibrating voltages and frequencies at the factory and preprogramming a series of defined voltage and frequency steps that all CPUs must achieve, AVFS dynamically measures and delivers the voltage required for each individual CPU to hit its desired clock frequencies. This allows for finer-grained power management across the CPU, improving both performance and power efficiency across a range of targets.

The 110-degree junction temperature is not evidence of a problem or a sudden issue with AMD graphics cards.SEEAMAZON_ET_135 See Amazon ET commerce AMD now measures its GPU temperature in new locations and reports additional data points that capture this information because it adopted more sophisticated measuring methods. Arguing that the company should be penalized for reporting data more accurately is akin to arguing that manufacturers ought to hide data because they’re afraid some customers won’t understand it or put it in the proper context.

AMD provides a pair of graphs to illustrate the difference between its Vega 64 and earlier measurement system and how it calibrates voltage on the 5700 XT today. The old discrete state method is shown below:

Vega64-DPM-States

Now, compare that against the frequency/voltage curve for the 5700 XT.

Fine-Grained-DPM

The 5700 XT is designed to continue boosting performance until it hits its thermal junction threshold. From the company’s blog post:

Paired with this array of sensors is the ability to identify the ‘hotspot’ across the GPU die. Instead of setting a conservative, ‘worst case’ throttling temperature for the entire die, the Radeon RX 5700 series GPUs will continue to opportunistically and aggressively ramp clocks until any one of the many available sensors hits the ‘hotspot’ or ‘Junction’ temperature of 110 degrees Celsius. Operating at up to 110C Junction Temperature during typical gaming usage is expected and within spec. This enables the Radeon RX 5700 series GPUs to offer much higher performance and clocks out of the box, while maintaining acoustic and reliability targets.

There’s a certain knee-jerk “I don’t want 110-degree anything in my case!” reaction from enthusiasts that’s both perfectly understandable and somewhat misguided. There’s an unconscious underlying assumption that 110 degrees Celsius represents a dangerous temperature (it doesn’t) or an extremely loud cooler. The 5700 XT and 5700 are much quieter than Vega 64, but if that’s still too loud, third-party cards are starting to hit the market. Companies like Asus were able to build coolers that handled the R9 290X beautifully, so the 5700 XT should be tamable as well.

Higher temperatures are partially an artifact of better measurement. They’re also a reality of advanced silicon manufacturing nodes. Our ability to pack transistors closer together has outstripped our ability to reduce their power consumption by cutting operating voltages. As a result, increasing transistor density increases hot spot formation and higher peak temperatures. AVFS helps mitigate this tendency by ensuring that operating voltage is precisely mapped to frequency, but it can’t fix the fact that AMD has packed more transistors into a smaller space, leading to higher thermal density.

Higher temperatures are not an intrinsic reason to be concerned about a product provided the manufacturer certifies that this is expected behavior. When I got into computing, a CPU temperature of 50 C (measured via in-socket thermistor) was considered extremely high. Today, Intel and AMD build silicon that can operate reliably at 95C or above for years at a time.

Now Read:




10 minutes mail – Also known by names like : 10minemail, 10minutemail, 10mins email, mail 10 minutes, 10 minute e-mail, 10min mail, 10minute email or 10 minute temporary email. 10 minute email address is a disposable temporary email that self-destructed after a 10 minutes. https://tempemail.co/– is most advanced throwaway email service that helps you avoid spam and stay safe. Try tempemail and you can view content, post comments or download something

No, AMD Hasn’t Quit Making Reference 5700 and 5700 XT GPUs


This site may earn affiliate commissions from the links on this page. Terms of use.

There’s an odd rumor going around that AMD has killed off its reference RX 5700 and RX 5700 XT GPU designs, or that it intends to do so once AIB’s custom cards are in-market. It started with French site Cowcatland, which ran the following headline:

CowCotLand

The translation of that headline states that AMD’s reference GPUs for the 5700 and 5700 XTSEEAMAZON_ET_135 See Amazon ET commerce have both reached EOL status only five weeks post-launch. It’s not true. According to AMD, the goal and point here are not to compete with AIB partners. “We expect there will continue to be strong supply of Radeon RX 5700 series graphics cards in the market, with multiple designs starting to arrive from our AIB partners,” AMD said. “As is standard practice, once the inventory of the AMD reference cards has been sold, AMD will continue to support new partner designs with Radeon RX 5700 series reference design kit.”

AMD provides reference designs for AIBs that want to speed cards to market without designing their own reference coolers or graphics boards. Early boards are typically based on these reference products. The delay between AIB shipments and reference card availability can be relatively short or can lag for some weeks. Some fans are unhappy that it’s been five weeks at this point without AIB designs, though we’ve seen this happen with Nvidia launches as well in the past. AMD isn’t killing off its reference cards, and they’ll still be manufactured going forward.

The enthusiast community isn’t particularly happy with the delay in blower cards or the fact that these cards are blowers, or the fact that the 5700 and 5700 XT remain noisier than equivalent Nvidia GPUs. The hope, therefore, is that dual or triaxial fan coolers will provide better acoustics than AMD’s default reference designs. This is, generally speaking, a pretty good bet.

Having tested the 5700, 5700 XT, Vega 64, Radeon VII, and an associated mixture of 2060, 2070, 2080, and 2080 Ti parts (both made by Nvidia and not), I’d say that honestly, the battle over a blower versus an open-air cooler can be a little inflated. Thermally, there’s an obvious difference between the two solutions (blowers exhaust hot air, while open-air coolers just move it around inside the chassis). What that difference means for your system depends a lot on what your system preconditions are. Open-air coolers can offer higher-performance in roomy cases with good airflow, while blowers provide more consistent results. The relative volume of the two solutions depends on their cooler design. A blower can be louder than an open-air cooler or vice-versa. The 5700 XT (a blower) is far quieter than Vega 64 (another blower). Vega 64 and the Radeon VII (an open-air design) have very similar noise profiles.

One interesting thing about reviews of Navi, however, is the degree to which the noise measurements from different review sites diverge. Anandtech, for example, reports that the 5700 XT is a 54dB(A) solution compared with 61dB for the Radeon Vega 64.

Image by Anandtech

This 54/61dB(A) solution seems to conform more closely to my own subjective experience of using the Radeon Vega 64, Radeon VII, 5700 XT, and associated Nvidia GPUs.SEEAMAZON_ET_135 See Amazon ET commerce The reason why I say this is because, to my own ear, the 5700 XT is vastly better than either the Radeon 64 or Radeon VII, both of which recall the Bad Old Days of loud GPUs like the R9 290X.

Other reviews, however, make very different claims:

Image by Guru3D

Guru3D claims that the Vega 64 and Radeon 5700 XT are identical in terms of db(A) and that the Radeon VII is significantly louder. Since distance from target obviously impacts noise measurements, I’m not concerned with the fact that Anandtech and Guru3D measure different levels of sound. What’s far more interesting is that one article shows Vega 64 and 5700 XT as comparable, while the other very much does not.

Image by TechPowerUp

TechPowerUp has a third distribution, with the 5700 XT and 5700 scoring identically and the Radeon VII below the Vega 64. Three well-regarded websites for tech reviews, three distinct results. Based on my own subjective experience, the one that “looks” the most correct is Anandtech’s — but noise measurements are going to be impacted by a number of factors, including relative levels of background noise, case-open testing versus case-closed, the distance from the target, and the equipment used to perform the test. It’s also possible that individual GPU variation is at work here as well.

In my own opinion, the 5700 and 5700 XT are firmly on the “Quiet enough” side of the “Is this GPU quiet enough to use or not?” It is not as quiet as the RTX 2060 or 2070 that we tested for the same review. It is considerably quieter than the Radeon VII or Vega 64. I have been known to wear earplugs when testing both of those cards in case-open configuration to avoid hearing damage, though the fact that I already have fan-related hearing damage in my left ear has also made me paranoid of harming it further. I’ve used a Vega 64 in my own system and disliked how noisy it was for gaming without headphones. The Radeon 5700 XT doesn’t cause the same issue.

Radeon AIB cards have often been quieter than the reference designs and so it’s likely this will continue to be the case. Whether these cards will offer reasonable values for the money is something we’ll check when they hit the market in larger quantities. Reference card designs will continue to exist alongside these newer cards as well.

Now Read:




10 minutes mail – Also known by names like : 10minemail, 10minutemail, 10mins email, mail 10 minutes, 10 minute e-mail, 10min mail, 10minute email or 10 minute temporary email. 10 minute email address is a disposable temporary email that self-destructed after a 10 minutes. https://tempemail.co/– is most advanced throwaway email service that helps you avoid spam and stay safe. Try tempemail and you can view content, post comments or download something

Ice Lake Benchmarks Paint a Complex Picture for Intel’s Latest CPU


This site may earn affiliate commissions from the links on this page. Terms of use.

Intel dropped a lot of Ice Lake news today, including an embargo lift on performance data concerning the new CPU. ExtremeTech was not aware that Intel had held a testing event in which reporters from various sites were invited to benchmark Ice Lake under controlled conditions and did not include this data in our initial coverage earlier today. We have reached out to Intel to clarify the situation, given that we were on hand at Architecture Day last winter to report on the initial Ice Lake CPU architecture and have covered Intel’s foundry research and developments for the past eight years.

Now that we actually know what CPU performance factually looks like, we’ve got a much better basis for discussing it relative to Intel’s Whiskey Lake. Our sister site PCMag has done a thorough comparison of Ice Lake against Whiskey Lake, with the Core i7-8565U represented in multiple form factors and systems from different OEMs. That’s actually incredibly useful because it shows just how large the gap between laptops can be, and how important proper testing (and thermals) are.

We’re going to excerpt benchmarks from the PCMag article and highly recommend you read the full story for that publication’s in-depth analysis. Let’s start with Cinebench R15:

657639-intel-ice-lake-cpu-tests-cinebench-r15

Right off the bat, we can see that Ice Lake has some issues in a 15W envelope. The fact that the CPU’s single-threaded performance improves by 1.22x when given room to breathe in a 25W design is evidence that CPU power consumption is throttling the core badly. There’s only a 5 percent spread between the Core i7-8565U machines as far as single-thread is concerned. When we move to multi-threading, giving the CPU 1.66x more thermal headroom results in a 1.33x improvement in performance. Comparing 15W with 15W, the older Intel CPUs are all faster, particularly the HP Envy 13.SEEAMAZON_ET_135 See Amazon ET commerce

In the 25W configuration, ICL wins the benchmark overall but is only 5 percent faster than the HP Envy 13. The Ryzen 5 2500U is outperformed (PCMag did not have a Ryzen 7 to compare against or an updated 3000-series APU in a mobile system).

POV-RAY shows some very interesting performance figures, in part because they’re completely different from the Handbrake distribution. The HP Envy 13, which was the fastest Core i7-8565U in Cinebench R15, is the slowest system in Handbrake (apart from the Pentium Gold, which doesn’t really count for our purposes). The Zenbook 13 is a whopping 20 percent faster than the 15W ICL testbed, though that system’s performance is on par with the Spectre X360 and Envy 13. Giving the CPU 25W to play with instead of 15W improves performance by about 24 percent, allowing ICL to beat past its rivals.

There are other results for the CPU-side available at PCMag and I’d look at them for a more complete picture. What we see in aggregate is that a 15W power envelope is a tight fit for the 10th Gen CPU family. Sometimes ICL is a bit faster than the 14nm Whiskey Lake CPUs, sometimes it’s slower, but we don’t see much evidence of improvement in the lower power envelope.

At the same time, however, we also see substantial variation in 14nm Core i7-8565U results. This isn’t surprising; Intel started giving OEMs more freedom to design SKUs back when Core M debuted, but all systems are not created equal. Certain laptops may be noticeably faster than others in certain circumstances. We recently talked about how increased variation in silicon performance explains many of AMD’s decisions around 7nm and the company’s Ryzen 7 products. This is a variation of a decidedly different sort, but that’s actually the point. Silicon companies have begun to design around variance in many ways because simply annihilating it has proven either prohibitively expensive or downright impossible.

That addresses the CPU component of Ice Lake. What about the GPU? Here, the news is much more positive.

In Rise of the Tomb Raider Low, ICL can maintain 40fps at 1366×768 and 26fps at 1920×1080. Interestingly, giving the system more headroom for power took the score down, not up at 768p and held it constant in 1080p. AMD’s lower-end Vega 8 does not compete well here, and while Vega 11 would provide some additional GPU headroom, it’s unlikely to completely close the gap. Only the MX150 and MX250-equipped laptops, with Nvidia GPUs,SEEAMAZON_ET_135 See Amazon ET commerce beat out Intel’s integrated graphics.

Tom Clancy’s Rainbow Six: Siege is impressive, with equal performance between the 15W and 25W CPUs. Again, only the MX150 and MX250 exceed Intel’s integrated performance. AMD’s Vega 8 turns in playable performance at 1366×768 but doesn’t meet the minimum 30fps threshold we consider minimum for 1080p gaming.

All of the GPU figures basically follow this pattern. You’ll see AMD hold its ground better in some than others, but Intel is ahead on the whole. Vega 11 would improve these results, but likely not by enough to change the outcome in most games.

Implications and Conclusion

In our earlier coverage written today, I implied that one reason for Intel’s lower CPU clocks might be because Intel used a larger amount of TDP to provide GPU performance. While there’s probably some truth to this, the 15W-25W performance pattern is different for CPUs compared with GPUs. Moving from 15W to 25W almost always improves CPU performance. Moving from 15W to 25W improves synthetic GPU benchmark performance on ICL, but has a weaker impact on actual games. Only World of Tanks enCore appears to respond strongly to the additional TDP headroom, suggesting that in most cases, that additional wattage isn’t going to the GPU — it’s being used to accelerate the CPU.

Gains for Ice Lake relative to Whiskey Lake are fairly anemic, though this can vary dramatically depending on which Whiskey Lake system you own now. When there’s a 10-15 percent variance between different systems equipped with the same processor, that’s obviously going to impact how ICL compares. Overall, we’d say Ice Lake is comparable to Whiskey Lake — sometimes faster, sometimes slower, but rarely dramatically distinguishing itself one way or the other.

The GPU improvements, on the other hand, are enormous. Assuming that the Ryzen 7 3700U and 3500USEEAMAZON_ET_135 See Amazon ET commerce are a relatively modest improvement on their predecessors, AMD will need to have 7nm APUs in-market to take ICL on. We have no timeline on when that may happen. Of course, AMD is currently focused on the desktop and server spaces, which means we don’t even know when Intel’s 10nm silicon will face off against AMD’s 7nm in-market.

The third pillar is power consumption and battery life, and we don’t know yet how ICL compares on these metrics; Intel forbid testing the sample laptop for such things. Right now, Ice Lake delivers massive improvements in one area, settles for small gains to small losses in another, and offers an unknown level of improvement in the third. Gamers who want some ability to play on thin-and-lights should be the major beneficiaries of the improvements we’ve seen thus far. If this performance holds, AMD will either need to hit back at Intel on 7nm or see its long domination of the integrated mobile GPU market finally fall — which isn’t honestly a sentence I used to think I’d ever type.

Now Read:




10 minutes mail – Also known by names like : 10minemail, 10minutemail, 10mins email, mail 10 minutes, 10 minute e-mail, 10min mail, 10minute email or 10 minute temporary email. 10 minute email address is a disposable temporary email that self-destructed after a 10 minutes. https://tempemail.co/– is most advanced throwaway email service that helps you avoid spam and stay safe. Try tempemail and you can view content, post comments or download something